30 April 2009 in Journal of Physics D: Applied Physics
3 March 2023 in IEEE Transactions on Multimedia
Open set action recognition (OSAR) is a rising research domain that simultaneously identifies all videos from known classes and rejects videos from unknown classes. Existing methods rarely consider the open set data distribution and the spatial-temporal relations of video subsequence. Recently proposed Capsule Network (CapsNet) has shown robust performance in many fields, especially image recognition. However, the current CapsNet has not been directly applied to the OSAR task since it cannot explicitly consider the data distribution of known and unknown classes along with the spatial-temporal relations for videos. This paper proposes the Spatial-Temporal Exclusive Capsule Network (STE-CapsNet) to solve the problems in the OSAR task. The STE-CapsNet designs the temporal-spatial routing mechanism to jointly capture the spatial-temporal information of the videos. Furthermore, the exclusive capsules are learned with dot product routing mechanism to limit the data distribution of closed set and open set and reduce the open set risk for OSAR. Extensive experimental results demonstrate that our proposed approach performs favorably compared with state-of-the-art methods on three standard datasets, which verifies its effectiveness and generalization ability.
15 August 2000 in Journal of Applied Physics
A combined study of the crystalline structure, the chemical interaction, and diffusion processes of the substrate/layer interface of amorphous-carbon hard coatings is presented. The structure of the coatings and their gradient layer interface to a chromium buffer layer has been investigated on two substrates [Si(100) and tool steel] using x-ray diffraction (XRD). Chemical interaction and diffusion processes at the interfaces and within the layers were analyzed by Auger electron spectroscopy and x-ray photoemission spectroscopy depth profiles. The chromium buffer layer revealed similar textured structure on both substrates. The subsequent gradient layer was determined (within XRD limits) to be amorphous and composed of an amorphous-carbon and chromium-carbide composite. The chromium carbide maintains the same stoichiometry (Cr3C2), regardless of the gradually changing chromium content. No large-scale interdiffusion was measured, either between or within the layers.
4 November 2023 in Knowledge-Based Systems